JP2007278168A - Fuel injection control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a fuel injection control device capable of keeping control properties of output of an internal combustion engine high even in a case that a plurality of times of fuel injection are performed in one combustion cycle. <P>SOLUTION: Predicted value of concentration of oxygen discharged from a combustion chamber is calculated based on demand injection quantity, opening of an EGR valve and air quantity in a step S22. Then, slippage quantity between the demand injection quantity and actual injection quantity is calculated based on difference between the predicted value and a detection value in a step S 26. Then, a learning value is calculated by dividing the slippage quantity by number of times of injection in multiple stage injection control in a step S28. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a fuel injection control device that controls the output of an internal combustion engine by multistage injection in which fuel is injected a plurality of times within one combustion cycle.

  In order to control the exhaust characteristics discharged to the exhaust system of a diesel engine, it is well known to provide an oxygen concentration sensor in the exhaust system. Here, it is intended to maintain good exhaust characteristics by feedback control based on the difference between the oxygen concentration detected by the oxygen concentration sensor and the target oxygen concentration. However, after exhaust gas is discharged into the exhaust system, there is a delay in response, particularly during a transition in which the operating state of the diesel engine changes, until the oxygen concentration in the exhaust gas is detected by the oxygen concentration sensor. For this reason, a deviation occurs between the oxygen concentration in the combustion chamber of the diesel engine and the oxygen concentration detected by the oxygen sensor, and as a result, the controllability of the exhaust characteristics may be reduced.

  Therefore, conventionally, for example, as seen in Patent Document 1 below, there is also a control device that predicts the oxygen concentration in the exhaust of a diesel engine and performs feedback control based on the difference between the predicted oxygen concentration and the target oxygen concentration. Proposed. According to this control device, the problem of the response delay can be avoided, and as a result, the controllability of the exhaust characteristics can be kept high.

  Further, in the above control device, in the steady operation state of the diesel engine, the predicted value shifts based on the difference between the predicted oxygen concentration and the detected value in each of the plurality of regions determined by the fuel injection amount and the rotational speed. It has also been proposed to learn quantities. Thus, for example, even if a deviation occurs in the predicted value because the injection characteristic of the fuel injection valve deviates from the reference injection characteristic assumed when predicting the oxygen concentration, the prediction is performed while compensating for the deviation amount. Is possible.

  By the way, in a diesel engine, usually, multistage injection control is performed in which fuel injection is performed a plurality of times within one combustion cycle in a single cylinder. In this case, when the injection characteristic of the fuel injection valve is deviated from the reference characteristic, the deviation amount of the predicted value of the oxygen concentration in each combustion cycle differs depending on the difference in the number of injections. On the other hand, the number of injections of multistage injection is basically determined by the rotational speed of the output shaft of the diesel engine and the required injection amount. In this case, as in the above control device, it is possible to compensate for the predicted shift amount of the oxygen concentration due to the difference in the injection stage by learning the shift amount for each of a plurality of regions determined by the fuel injection amount and the rotation speed. it can.

  However, in recent years, the number of injections has been changed depending on various factors such as whether or not the diesel engine has been warmed up due to demands for further improvement of exhaust characteristics and noise suppression. In this case, since the difference in the injection amount due to the difference in the number of injections cannot be reflected by learning the amount of deviation for each region, it is possible to appropriately compensate for the amount of deviation in the predicted value of the oxygen concentration. Can not. On the other hand, it is also conceivable that a region is defined using all the parameters for determining the number of injections, and a deviation amount is learned for each region. However, in this case, the number of areas becomes enormous, and learning opportunities in each area are remarkably reduced, so that it cannot be practically used.

In the fuel injection control device that controls the output of the internal combustion engine by multi-stage injection, not only the control device described above, the total deviation amount of the injection amount due to the deviation in the injection characteristics varies depending on the injection stage. For this reason, such a situation that the controllability of the output of the internal combustion engine is lowered is also common.
JP 2002-327634 A

  The present invention has been made in order to solve the above-described problems, and the object thereof is to maintain high controllability of the output of an internal combustion engine even when a plurality of fuel injections are performed in one combustion cycle. An object of the present invention is to provide a fuel injection control device that can be used.

  Hereinafter, means for solving the above-described problems and the operation and effects thereof will be described.

  According to a first aspect of the present invention, a means for calculating a predicted value of the oxygen concentration in the exhaust gas of the internal combustion engine accompanying the fuel injection control, and a difference between the detected value of the oxygen concentration and the predicted value are calculated by the multistage injection. Learning means for learning a deviation amount of the injection characteristic of the fuel injection valve of the internal combustion engine by converting into a difference due to one injection.

  In the above configuration, the difference between the detected value and the predicted value of the oxygen concentration is considered to be caused by the deviation of the injection characteristic of the fuel injection valve from the reference characteristic assumed in the prediction of the oxygen concentration. However, the difference between the detected value and the predicted value depends not only on the deviation in the injection characteristics but also on the number of injections. In this regard, in the above configuration, the difference between the detected value and the predicted value is converted into the difference due to the injection per multistage injection, thereby removing the influence of the number of injections and learning the deviation amount of the injection characteristics. it can. For this reason, the controllability of the output can be maintained high by correcting the operation amount of the actuator for controlling the output based on the learned deviation amount.

  According to a second aspect of the present invention, in the first aspect of the invention, the deviation in the injection characteristic is caused by a deviation in an injection period due to a deviation in a delay amount of an actual injection start timing with respect to a command value of the injection start timing. According to another aspect of the present invention, the learning means learns, as the shift amount, a value obtained by quantifying the shift in the injection period.

  In the above configuration, focusing on the fact that the injection period is shifted due to a shift in the actual injection start timing delay amount with respect to the command value of the injection start timing, the amount of shift in the injection period is quantified. Since this quantification can be performed as a quantification of an element having a correlation with an injection period deviation, such as an injection period deviation itself or an injection quantity deviation caused by an injection period deviation, an injection characteristic deviation amount The number of parameters used in quantifying can be reduced.

  According to a third aspect of the present invention, in the first or second aspect of the invention, the learning means is configured to perform the multistage injection based on a difference between the detected value and the predicted value and an amount of air to be burned in the internal combustion engine. And means for calculating a total deviation amount of the injection amount, and a calculation means for calculating the deviation amount of the injection characteristic by dividing the total deviation amount by the number of injections of the multistage injection. To do.

  In the above configuration, by multiplying the difference between the predicted value and the detected value by the amount of air to be used for combustion, the difference in the amount of oxygen per the amount of air can be calculated. This difference in oxygen amount is considered to be caused by a difference in fuel injection amount. For this reason, based on the amount of oxygen required to burn the unit amount of fuel, the difference in oxygen amount can be converted into the total deviation amount. Then, by dividing the total deviation amount by the number of injections, the deviation amount of the injection amount accompanying one injection can be calculated.

  According to a fourth aspect of the present invention, in the invention of the second aspect, the fuel injection valve has an inflection point at which the proportionality coefficient changes between the injection period and the injection amount. And the learning means calculates a total deviation amount of the injection amounts associated with the multistage injection based on a difference between the detected value and the predicted value and an air amount to be burned in the internal combustion engine. And a deviation amount of the injection characteristic from the total deviation amount based on the number of injections of the multistage injection and the inflection point on the assumption that the injection period is shifted by a deviation amount of the delay amount of the actual injection start timing. And calculating means for calculating.

  In the above configuration, by multiplying the difference between the predicted value and the detected value by the amount of air to be used for combustion, the difference in the amount of oxygen per the amount of air can be calculated. This difference in oxygen amount is considered to be caused by a difference in fuel injection amount. For this reason, based on the amount of oxygen required to burn the unit amount of fuel, the difference in oxygen amount can be converted into the total deviation amount.

  On the other hand, when the injection period deviates by the actual injection start timing, the entire straight line that is determined by the injection period and the injection amount and indicates the reference injection characteristic is shifted in the injection period direction by the above-described deviation. The straight line obtained in this way represents the actual injection characteristics. For this reason, the deviation amount of the fuel injection due to the injection of each injection stage can be uniquely determined by the number of injections, the inflection point, and the total deviation amount.

  According to a fifth aspect of the present invention, in the invention according to any one of the first to fourth aspects, the learning means defines the amount of deviation of the injection characteristic by the pressure of the fuel supplied to the fuel injection valve. Learning is performed for each of a plurality of regions, and further includes storage means for storing a deviation amount of the injection characteristic learned for each of the plurality of regions.

  The amount of deviation in the injection characteristics tends to depend significantly on the pressure of the fuel supplied to the fuel injection valve. In this regard, in the above configuration, since it is possible to learn the fuel pressure dependence of the amount of deviation of the injection characteristic by learning the amount of deviation of the injection characteristic for each of a plurality of regions defined by the fuel pressure, the deviation of the injection characteristic The amount can be learned with high accuracy.

  According to a sixth aspect of the present invention, in the invention according to any one of the first to fifth aspects, the learning means determines the amount of deviation in the injection characteristics, the pressure of fuel supplied to the fuel injection valve, and the internal combustion engine. And learning means for each of the plurality of regions defined by the rotational speed of the output shaft, and further comprising storage means for storing the deviation amount of the injection characteristic learned for each of the plurality of regions. .

  When the deviation amount of the injection characteristic is learned according to the difference between the predicted value and the detected value of the oxygen concentration, there is a risk of being influenced by a detection error of the means for detecting the oxygen concentration, a detection error of the intake air amount, etc. . In this case, since the detection error is actually reflected in what is learned as the deviation amount of the injection characteristic, the deviation amount may vary depending on the operating state of the internal combustion engine. In order to avoid the fluctuation in the learning result of the deviation amount due to the detection error, the deviation amount is learned for each region divided by the rotation speed and the injection amount that cause the fluctuation of the oxygen concentration and the intake air amount. It is desirable to do. On the other hand, the amount of deviation in the injection characteristics tends to depend significantly on the fuel pressure. For this reason, it is desirable that the deviation amount of the injection characteristic is learned for each of a plurality of regions determined by the fuel pressure. Since the fuel pressure is determined according to the injection amount and the rotation speed, if the deviation amount is learned for each region divided by the rotation speed and the fuel pressure, the fuel pressure is determined according to the rotation speed and the injection amount. In addition, it is possible to suppress the fluctuation due to the detection error and to learn the fuel pressure dependence of the deviation amount of the injection characteristic.

  The invention according to claim 7 is the invention according to any one of claims 1 to 6, wherein target value calculation means for calculating a target value of the oxygen concentration in the exhaust gas and feedback control of the predicted value to the target value. Therefore, it further comprises an operating means for operating an actuator for controlling the oxygen concentration in the exhaust, and the operating means uses the deviation amount learned by the learning means as the predicted value used for the feedback control. It is characterized by using the predicted value calculated by using.

  In the above configuration, in order to feedback control the predicted value to the target value, compared with the case where the detected value of the oxygen concentration in the exhaust gas is feedback controlled to the target value, the exhaust gas is discharged from the combustion chamber during the transient operation of the internal combustion engine. Therefore, it is possible to suitably avoid a decrease in controllability due to a difference between the actual oxygen concentration in the exhaust gas and the detected value. In addition, by calculating the predicted value used for the feedback control using the deviation amount of the injection characteristic learned by the learning means, the deviation of the injection characteristic of the fuel injection valve can be compensated.

  The invention according to claim 8 is the invention according to claim 7, wherein the internal combustion engine adjusts a flow area of the exhaust gas recirculation passage for recirculating the exhaust gas discharged to the exhaust system to the intake system and the exhaust gas recirculation passage. And an EGR valve, wherein the operation means operates an opening of the EGR valve so as to feedback control the predicted value to the target value.

  In the above configuration, the oxygen concentration can be feedback controlled to the target value by operating the opening of the EGR valve. In particular, in order to correct the deviation of the injection characteristic by the recirculation amount of the exhaust gas, even if the detection error of the detecting means is reflected in the learning of the deviation, the detection error is caused in comparison with the case of correcting the injection amount. It is difficult to cause unintended fluctuations in output torque.

  The invention according to claim 9 is the invention according to any one of claims 1 to 6, further comprising an opening / closing operation means for opening / closing the fuel injection valve based on a command value of an injection amount for the fuel injection valve, The opening / closing operation means corrects the setting according to the deviation amount learned by the learning means when setting the operation amount of the fuel injection valve according to the command value.

  In the above configuration, when setting the operation amount of the fuel injection valve according to the command value, by correcting the setting according to the deviation amount, the actual injection amount and the command value regardless of the deviation in the injection characteristics. Can be matched with high accuracy.

(First embodiment)
Hereinafter, a first embodiment in which a fuel injection control device according to the present invention is applied to a fuel injection control device of a common rail diesel engine will be described with reference to the drawings.

  FIG. 1 shows the overall configuration of the engine system according to the present embodiment.

  As shown in the drawing, a throttle valve 14 is provided upstream of the intake passage 12 of the diesel engine 10. The intake passage 12 communicates with the combustion chamber 18 by opening the intake valve 16. The tip of the fuel injection valve 40 projects into the combustion chamber 18, and fuel is supplied by the fuel injection valve 40. As a result, combustion of fuel occurs in the combustion chamber 18, and this combustion energy is converted into kinetic energy of the piston 20. The combustion chamber 18 communicates with the exhaust passage 24 by opening the exhaust valve 22. An exhaust gas purification device 26 is provided in the exhaust passage 24. Further, the exhaust passage 24 and the intake passage 12 can be communicated with each other by an exhaust recirculation passage 28 in order to recirculate the exhaust discharged to the exhaust passage 24 to the intake passage 12. The flow area of the exhaust gas recirculation passage 28 is adjusted by operating the opening degree of the EGR valve 30 by the valve actuator 32.

  The fuel injection valve 40 injects and supplies high pressure fuel supplied from a common rail (not shown) to the combustion chamber 18 of the diesel engine 10. Specifically, the fuel injection valve 40 is provided with a cylindrical needle storage portion 42 at the tip thereof. The needle storage portion 42 stores a nozzle needle 44 that can be displaced in the axial direction. The nozzle needle 44 is seated on an annular needle seat portion 46 formed at the distal end portion of the fuel injection valve 40, thereby blocking the needle storage portion 42 from the outside (combustion chamber 18), and from the needle seat portion 46. By separating, the needle storage portion 42 communicates with the outside. Further, high pressure fuel is supplied from the common rail to the needle storage portion 42 via the high pressure fuel passage 48.

  The back side of the nozzle needle 34 (the side opposite to the side facing the needle seat portion 46) faces the back pressure chamber 50. High pressure fuel is supplied to the back pressure chamber 50 from the common rail via the high pressure fuel passage 48. Further, a needle spring 52 is provided at an intermediate portion of the nozzle needle 44, and the nozzle needle 44 is pushed toward the distal end side of the fuel injection valve 40 by the needle spring 52.

  On the other hand, the back pressure chamber 50 can communicate with the low pressure fuel passage 56 via the orifice 54. The low pressure fuel passage 56 is connected to a fuel tank. The back pressure chamber 50 and the low pressure fuel passage 56 are communicated and blocked by a valve body 58. That is, the orifice 54 communicating the back pressure chamber 50 and the low pressure fuel passage 56 is closed by the valve body 58, whereby the back pressure chamber 50 and the low pressure fuel passage 56 are blocked, while the orifice 54 is opened. Thus, the back pressure chamber 50 and the low pressure fuel passage 56 are communicated with each other.

  The valve body 58 is pushed toward the distal end side of the fuel injection valve 40 by a valve spring 60. Further, the valve body 58 can be displaced to the rear side of the fuel injection valve 40 by being attracted by the electromagnetic force of the electromagnetic solenoid 62.

  In such a configuration, when the electromagnetic solenoid 62 is not energized and the attraction force by the electromagnetic solenoid 62 is not working, the valve body 58 closes the orifice 54 by the force of the valve spring 60. On the other hand, the nozzle needle 44 is pushed toward the distal end side of the fuel injection valve 40 by the needle spring 52 and is in a state of being seated on the needle seat portion 46 (the fuel injection valve 40 is closed).

  Here, when the electromagnetic solenoid 62 is energized, the valve body 58 is displaced to the rear side of the fuel injection valve 40 by the suction force of the electromagnetic solenoid 62 and opens the orifice 54. As a result, the high pressure fuel in the back pressure chamber 50 flows out to the low pressure fuel passage 56 via the orifice 54. For this reason, the pressure applied to the nozzle needle 44 by the high-pressure fuel in the back pressure chamber 50 is smaller than the pressure applied to the nozzle needle 44 by the high-pressure fuel in the needle housing portion 42. When the force due to this pressure difference becomes larger than the force by which the needle spring 52 pushes the nozzle needle 44 toward the tip of the fuel injection valve 40, the nozzle needle 44 is separated from the needle seat portion 46 (the fuel injection valve 40). Open state).

  The engine system further includes an air flow meter 70 for detecting the amount of intake air taken into the intake passage 12, particularly the amount of intake air upstream of the throttle valve 14, and an intake pressure sensor for detecting the pressure in the intake passage 12. 72, an intake air temperature sensor 74 that detects the temperature of the intake air, an air-fuel ratio sensor 76 that detects the oxygen concentration in the exhaust downstream of the exhaust purification device 26, a crank angle sensor 78 that detects the rotation angle of the output shaft of the diesel engine 10, and a common rail Various sensors for detecting the operating state of the diesel engine 10 are provided, such as a fuel pressure sensor 80 for detecting the fuel pressure in the engine and a water temperature sensor 82 for detecting the temperature of the cooling water of the diesel engine 10. The engine system also includes an accelerator sensor 84 that detects the amount of operation of the accelerator pedal.

  The engine system further includes an electronic control unit (ECU 90). The ECU 90 includes a central processing unit and a constant memory holding memory 92, takes in the detection values of the various sensors, and controls the output (output torque, exhaust characteristics, etc.) of the diesel engine 10 based on these values. Here, the constant memory holding memory 92 is a backup RAM that is always in the power supply state regardless of the state of the start switch (ignition switch) of the ECU 90, or a non-volatile memory that holds the stored contents regardless of the presence or absence of the power supply state ( This is a memory that holds the stored contents regardless of the state of the start switch of the ECU 90, such as an EEPROM.

  FIG. 2 shows a processing procedure of the fuel injection control for the output control. This process is repeatedly executed by the ECU 90, for example, at a predetermined cycle.

  In this series of processing, first, in step S10, the accelerator pedal operation is performed based on the operation amount of the accelerator pedal detected by the accelerator sensor 84 and the rotation speed of the output shaft of the diesel engine 10 based on the detection value of the crank angle sensor 78. The injection amount (requested injection amount) required to generate the output torque corresponding to the is calculated. FIG. 3 shows a map for calculating the injection amount from the accelerator pedal operation amount ACCP and the rotational speed. FIG. 3 also shows a map for setting the target value (target fuel pressure) of the fuel pressure in the common rail from the injection amount and the rotational speed calculated in this way. The target fuel pressure is set to a higher pressure as the rotational speed and the injection amount are larger according to the map shown in FIG.

  In subsequent step S12, the number of injection stages is set based on the required injection amount. This is a process for performing multi-stage injection control in which some of the pilot injection, main injection, and after injection are selected within one cycle of the combustion cycle, and these selected injections are performed. Here, in pilot injection, a very small amount of fuel is injected to promote the mixing of fuel and air immediately before ignition, and the ignition timing delay after main injection is shortened to generate nitrogen oxides (NOx). To reduce combustion noise and vibration. The main injection contributes to the generation of output torque of the diesel engine and has the maximum injection amount during multi-stage injection. After-injection recombusts particulate matter (PM). Here, for example, when the number of injection stages is “2”, one stage of pilot injection and one stage of main injection are performed, and when the number of injection stages is “4”, two stages of pilot injection and one stage of main injection are performed. And one-stage after injection.

  The setting of the number of injection stages is set according to not only the required injection amount but also the temperature of the cooling water detected by the water temperature sensor 82, for example.

  In the subsequent step S14, a command value (command injection start timing) for the injection start timing of each injection stage and a command value (command injection period) for the injection period of each injection stage are calculated. Here, the command injection period is map-calculated based on the fuel pressure detected by the fuel pressure sensor and the injection amount of the injection stage. In step S16, the fuel injection valve 40 is operated to inject fuel at each injection stage. In addition, when the process of step S16 is completed, this series of processes is once complete | finished.

  The map for setting the command injection period is generated on the assumption that the fuel injection valve 40 has a reference characteristic. However, the actual injection characteristics of the fuel injection valve 40 may deviate from the reference characteristics due to individual differences and changes over time. For this reason, in order to maintain high controllability of the output of the diesel engine 10, it is desirable to learn the deviation amount of the injection characteristics. However, when performing multi-stage injection as described above, since the output of the diesel engine 10 generated by fuel injection is affected by the average injection amount deviation due to multi-stage injection, the deviation amount can be learned. Have difficulty.

  That is, when the actual characteristics deviate from the reference injection characteristics indicated by the solid line in FIG. 4 as indicated by the alternate long and short dash line, the required injection quantity is changed to the four-stage injection of the command injection quantities Q11 to Q14. The amount of deviation in the injection amount of the multi-stage injection differs between when dividing and when dividing into two stages of command injection amounts Q11 and Q15. That is, in the case of the above-described four-stage injection, the difference between the command injection amounts Q11 to Q14 and the actual injection amounts Q21 to Q24 is the total deviation amount, whereas in the case of the two-stage injection, the command injection amount Q11. , Q15 and the actual injection amounts Q21, Q25 are the total deviation amount.

  Therefore, in the present embodiment, the oxygen concentration immediately after being discharged from the combustion chamber 18 of the diesel engine 10 (or the oxygen concentration in the combustion air in the combustion chamber 18) is predicted, and this predicted value and the value detected by the air-fuel ratio sensor 76 are calculated. Is converted into a difference resulting from the injection per multistage injection, thereby learning the amount of deviation in the injection characteristics. Here, the fact that the deviation of the injection characteristic of the fuel injection valve 40 has the properties shown in FIG. 4 is utilized. As shown in the figure, there is a proportional relationship between the injection amount and the injection period, and the deviation in the injection characteristic shifts the line defining the proportional relationship by a predetermined amount in the injection period direction. This is due to the following reason.

  That is, the deviation in the injection characteristic of the fuel injection valve 40 occurs as a deviation in the actual injection start timing with respect to the command injection start timing, as shown in FIG. FIG. 5A shows an energization signal for the fuel injection valve 40, and FIG. 5B shows an actual injection rate by the fuel injection valve 40. As shown in the drawing, energization to the fuel injection valve 40 is started at time t1 which is the command injection start timing, but the fuel injection valve 40 is actually opened and fuel injection is started later than this. Time t2 is reached. When the energization of the fuel injection valve 40 is stopped at time t3 when the command injection period elapses from the command injection start timing, the injection rate is decreased and the fuel injection is ended.

  Here, when the actual injection characteristic of the fuel injection valve 40 deviates from the reference injection characteristic, the actual injection start timing of the fuel injection valve 40 becomes the reference, as shown by a one-dot chain line in FIG. It will be shifted with respect to the thing. For this reason, the actual injection period changes by the deviation of the injection start timing, and as a result, the injection amount changes. Therefore, the actual injection characteristic is obtained by offsetting the reference characteristic line shown in FIG. 4 by “t2′−t2” in the injection period direction.

  According to the deviation in the injection characteristics shown in FIG. 4, the deviation amount of the injection amount is the same regardless of the length of the command injection period. Therefore, in the present embodiment, the difference in the injection amount per injection is calculated by dividing the difference between the actual amount of the total injection amount by the multistage injection and the amount by the reference characteristic by the number of injections.

  FIG. 6 shows the procedure of the learning process of the injection characteristic deviation amount according to the present embodiment. This process is repeatedly executed by the ECU 90, for example, at a predetermined cycle.

  In this series of processing, first, in step S20, it is determined whether or not a learning condition is satisfied. The learning conditions include that the state in which the amount of change in the rotational speed is less than or equal to a predetermined time continues for a specified time or that the state in which the amount of change in the injection amount is less than or equal to a predetermined time continues for a specified time. The operating state of the engine 10 may be a steady state.

  In the subsequent step S22, the oxygen concentration in the exhaust gas is predicted based on the required injection amount calculated in the processing of FIG. 2, the opening degree of the EGR valve 30, and the detection values of various sensors. Here, first, the oxygen concentration of the gas flowing into the combustion chamber 18 is calculated based on the detection value of the air flow meter 70, the detection value of the intake pressure sensor 72, the detection value of the intake air temperature sensor 74, the opening degree of the EGR valve 30, and the like. To do. As the calculation method, for example, the method described in Patent Document 1 may be used. When the oxygen concentration is calculated, the oxygen concentration in the exhaust gas is calculated by subtracting the oxygen amount consumed by the combustion of the required injection amount of fuel from the oxygen amount in the gas in the combustion chamber 18. This calculation method may also be the method described in Patent Document 1, for example.

  In the subsequent step S24, the detection value of the air-fuel ratio sensor 76 is acquired. In step S26, a deviation amount between the actually injected fuel amount and the required injection amount is calculated based on the deviation between the predicted value of the oxygen concentration in the process of step S22 and the detected value. That is, since the oxygen concentration necessary for burning this per unit injection amount can be obtained in advance, the difference between the predicted value and the detected value is multiplied by the amount of gas in the combustion chamber 18 to increase the oxygen concentration in the combustion chamber 18. By calculating this oxygen amount and dividing this oxygen amount by the required oxygen concentration, it is possible to calculate the total deviation amount of the injection amounts due to the multistage injection.

  In the subsequent step S28, a deviation amount (learning value) of the injection amount per injection is calculated. This can be calculated by gradually calculating the total injection amount deviation calculated in step S26 by the number of injections.

  When the learning value is calculated, the learning value is stored in the constant memory holding memory 92 shown in FIG. 1 in step S30. Specifically, as shown in FIG. 7, the learning value is learned and stored for each of a plurality of regions divided by the rotational speed of the output shaft of the diesel engine 10 and the fuel pressure in the common rail. This is due to the following reason.

  As described above, the deviation of the injection characteristic due to the deviation of the delay amount of the actual injection start timing with respect to the command injection start timing of the fuel injection valve 40 remarkably depends on the pressure of the fuel supplied to the fuel injection valve 40. To do. For this reason, it is desirable that the deviation amount of the injection characteristic is learned separately for each region divided by the fuel pressure. This is because, if the detected shift amount of the oxygen concentration is uniquely determined by the shift amount of the injection characteristic of the fuel injection valve 40, the learning value learned separately for each region divided for each fuel pressure is used. This means that the difference between the predicted value and the detected value of the oxygen concentration can be compensated with high accuracy.

  In practice, however, the cause of the difference between the predicted value and the detected value is predominantly due to the difference in the injection characteristics of the fuel injection valve 40, but detection of the sensor used to calculate the predicted value. Error is included. On the other hand, as shown in FIG. 3 above, even if the fuel pressure is constant, the rotational speed and the injection amount can take various values. For this reason, the intake air amount detected by the air flow meter 70 and the air-fuel ratio sensor The oxygen concentration detected by 76 can take various values. Therefore, if the learning value is stored for each of the plurality of regions divided by the fuel pressure, the magnitude of these detection errors varies due to the difference in rotational speed, injection amount, etc. even during learning at the same fuel pressure. As a result, the learning value may not converge.

  In order to manage the detection error by the sensor, it is desirable to learn the learning value for each region divided by the injection amount and the rotation speed. On the other hand, as described above, in order to manage the deviation amount of the injection characteristic of the fuel injection valve 40, it is desirable to learn the learning value for each region divided by the fuel pressure. Further, the fuel pressure is determined by the rotational speed and the injection amount as shown in FIG. This means that the injection amount is determined from the fuel pressure and the rotational speed. Therefore, if the learning value is learned for each region divided by the rotation speed and the fuel pressure, the detection error by the sensor can be managed. That is, in this way, the learning value detected in each region includes a specific amount of the region as a detection error by the sensor in addition to the deviation in the injection characteristics. For this reason, it is possible to avoid the learning value from fluctuating due to a change in the operating state of the diesel engine 10.

  When a negative determination is made in step S20 of FIG. 6 or when the process of step S30 is completed, this series of processes is temporarily terminated.

  FIG. 8 shows a processing procedure for correcting output control of the diesel engine 10 using the learning value in the present embodiment. This process is repeatedly executed by the ECU 90, for example, at a predetermined cycle.

  In this series of processes, first, in step S40, the basic value of the opening degree of the EGR valve 30 is set based on the operating state of the diesel engine 10. Here, for example, the basic value may be calculated by using a two-dimensional map that determines the basic value from the rotation speed and the injection amount. In the subsequent step S42, a target value of oxygen concentration (target oxygen concentration) is calculated based on the operating state of the diesel engine 10. Here, for example, a map calculation of the target oxygen concentration may be performed using a two-dimensional map that determines the target oxygen concentration from the rotation speed and the injection amount. Subsequently, in step S44, the predicted value of the oxygen concentration is calculated based on the learned value stored in the map shown in FIG. Here, the predicted value may be calculated by performing the process in step S22 of FIG. 6 using the corrected required injection amount with the learned value. That is, when the required injection amount is divided into, for example, the four injection amounts of the command injection amounts Q11 to Q14 shown in FIG. 4, the required injection amount is increased and corrected by a value that is four times the learning value. What is necessary is just to perform the said process. Further, when the required injection amount is divided into two injection amounts, that is, the command injection amounts Q11 and Q15, the required injection amount may be increased and corrected by a value obtained by doubling the learning value.

  In the subsequent step S46, a feedback correction amount based on the difference between the target oxygen concentration and the predicted value is calculated. Here, the predicted value is used as an accurate representation of the oxygen concentration in the combustion chamber 18 of the diesel engine 10. That is, since the oxygen concentration detected by the air-fuel ratio sensor 76 is accompanied by an error due to a response delay as compared with the oxygen concentration in the combustion chamber 18, here, correction is performed to feedback control the predicted value to the target oxygen concentration. Calculate the amount. In step S48, the basic value of the opening degree of the EGR valve 30 is corrected with the correction amount. In step S50, the actual oxygen concentration is feedback-controlled to the target oxygen concentration by operating the opening degree of the EGR valve 30 with the basic value corrected with the correction amount. Thereby, the opening degree of the EGR valve 30 is decreased to increase the oxygen concentration as the learning value increases due to the actual injection amount becoming larger than the command injection amount.

  As described above, in this embodiment, the deviation amount between the predicted value and the detected value of the oxygen concentration is learned as the deviation in the injection characteristics, and the deviation in the oxygen concentration is corrected by correcting the exhaust gas recirculation amount (EGR amount) by operating the EGR valve. Compensated. Thereby, even if there is a factor other than the deviation of the injection characteristic of the fuel injection valve 40 in the learning value, the oxygen concentration can be controlled with high accuracy. On the other hand, when the fuel injection amount by the fuel injection valve 40 is corrected, an unintended change in the output torque may be caused if the learning value includes the influence of the detection error of the sensor.

  According to the embodiment described in detail above, the following effects can be obtained.

  (1) By converting the difference between the predicted value and the detected value of the oxygen concentration in the exhaust of the diesel engine 10 associated with the fuel injection control into the difference due to the injection per multistage injection, the influence due to the number of injections is eliminated. Thus, the deviation amount of the injection characteristic of the fuel injection valve 40 of the diesel engine 10 can be learned.

  (2) The deviation in the injection period is quantified by assuming that the deviation in the injection characteristic is caused by the deviation in the injection period due to the deviation in the actual injection start timing from the command value of the injection start timing. Was learned as a deviation amount of the injection characteristic. Thereby, the number of parameters used when quantifying the deviation | shift amount of an injection characteristic can be reduced.

  (3) Based on the difference between the detected value and the predicted value of the oxygen concentration, the total deviation amount of the injection amount associated with the multistage injection is calculated, and this total deviation amount is divided by the number of injections of the multistage injection, thereby causing a deviation in the injection characteristics. The amount was calculated. Thereby, the amount of deviation of the injection amount accompanying one injection can be calculated as the amount of deviation of the injection characteristic.

  (4) By learning the deviation amount of the injection characteristic for each of a plurality of regions defined by the fuel pressure, it is possible to learn the fuel pressure dependency of the deviation amount of the injection characteristic.

  (5) By learning the deviation amount of the injection characteristic for each of a plurality of regions defined by the rotational speed and fuel pressure of the output shaft of the diesel engine 10, deviation amounts due to various detection errors according to the rotational speed and the injection amount Fluctuations can be suppressed.

  (6) When the opening degree of the EGR valve 30 is operated to feedback control the predicted value of the oxygen concentration to the target value, the predicted value calculated based on the learned deviation amount is used. Thereby, it is possible to suitably avoid a decrease in controllability due to a deviation between the actual oxygen concentration and the detected value during a transient operation of the diesel engine 10, and the oxygen concentration due to a deviation in the injection characteristics of the fuel injection valve 40. A decrease in controllability can be compensated.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  In the present embodiment, the fuel injection control itself is corrected based on the learned value of the deviation amount of the injection characteristic. FIG. 9 shows a processing procedure of fuel injection control according to the present embodiment. This process is repeatedly executed by the ECU 90, for example, at a predetermined cycle.

  In this series of processes, first, in step S60, a command injection amount calculated in association with the process of step S12 of FIG. 2 is acquired. In the subsequent step S62, the command injection amount is corrected by the learned value. Here, the learning value may be simply subtracted from the command injection amount. In the subsequent step S64, the corrected command injection amount is converted into a command injection period based on the corrected command injection amount and the fuel pressure. In step S66, the fuel injection valve 40 is operated in accordance with the command injection period.

  According to the present embodiment described above, the following effects can be obtained in addition to the effects (1) to (5) described above.

  (7) When setting the operation amount of the fuel injection valve 40, the setting was corrected according to the learned deviation amount of the injection characteristic. Thereby, the actual injection amount and the command value can be made to coincide with each other with high accuracy regardless of the deviation of the injection characteristics.

(Third embodiment)
Hereinafter, a third embodiment will be described with reference to the drawings, focusing on differences from the first and second embodiments.

  FIG. 10 shows the injection characteristics of the fuel injection valve 40 according to this embodiment. In the figure, a solid line indicates a characteristic that serves as a reference for the fuel injection valve 40. As shown in the figure, the injection characteristic of the fuel injection valve 40 is characterized by having an inflection point in which there is a proportional relationship between the injection period and the injection amount and the proportionality coefficient changes. In addition, an example in which the injection characteristic is deviated from the reference characteristic due to individual differences of the fuel injection valve 40 or a change with time is shown by a one-dot chain line. As shown in the figure, also in this case, the straight line relating the injection amount and the injection period is shifted in the injection period direction by the offset amount ΔTQ. For this reason, the value of the injection amount that defines the inflection point is unchanged, and the injection period that defines the inflection point is shifted by the offset amount ΔTQ.

  Therefore, in this embodiment, assuming that the straight line is shifted by the offset amount ΔTQ, the shift amount of the injection characteristic of the fuel injection valve 40 is calculated based on the number of injections and the inflection point. Specifically, there is a relationship of “ΔQu = α × ΔQd” between the deviation amount ΔQd at the injection amount smaller than the inflection point and the deviation amount ΔQu at the injection amount larger than the inflection point. The coefficient α can be determined by the slope of the straight line in the reference characteristics. For this reason, in this embodiment, this coefficient α is acquired in advance as a value unique to the fuel injection valve 40, and the deviation amount of the injection characteristic is calculated using this coefficient α.

That is, for example, in the four-stage fuel injection in which the command injection amount is Q11 to Q14, when the actual injection amount is Q21 to Q24, the deviation amount of the total injection amount by the four-stage injection is
3 × ΔQd + ΔQu = (3 + α) × ΔQd
It becomes.

  For this reason, the deviation amount ΔQd can be learned as the deviation amount of the injection characteristics. Thus, the output control mode of the diesel engine 10 can be corrected. Hereinafter, a case where the learning value according to the present embodiment is applied to the first embodiment and the second embodiment will be described.

<When applied to the first embodiment>
In this case, when calculating the predicted value of the oxygen concentration in step S44 of FIG. 8, the required injection amount may be corrected according to the command injection amount, the coefficient α, and the number of injections of each injection stage. For example, when the command injection amount is composed of Q11 to Q14 as shown in FIG. 10, the required injection amount may be corrected by “(3 + α) × ΔQd”.

<When applied to the second embodiment>
In this case, in step S62 of FIG. 9, each command injection amount may be corrected based on the coefficient α and the learned deviation amount ΔQd. For example, when the command injection amount is composed of Q11 to Q14 as shown in FIG. 10, the command injection amount Q1 to Q3 is corrected to decrease by the deviation amount ΔQd, and the command injection amount Q4 is “α × ΔQd”. Only a decrease correction.

  In the present embodiment described above, the effects (1), (2), (4) to (6) of the previous first embodiment and the effect (7) of the previous second embodiment are achieved. In addition, the following effects can be obtained.

  (8) The deviation of the fuel injection characteristic can be uniquely determined based on the deviation amount ΔQd on the minute injection side from the inflection point and the coefficient α.

(Other embodiments)
The above embodiments may be implemented with the following modifications.

  In each of the above embodiments, the deviation of the injection characteristic of the fuel injection valve 40 is caused by the deviation of the actual injection period by the deviation amount of the actual injection start timing with respect to the command injection start timing. Not exclusively. FIG. 11 (a) shows a command injection period, FIG. 11 (b) shows the lift amount of the nozzle needle 44, and FIG. 11 (c) shows each transition example in characteristics different from the above embodiments. In the drawing, as shown by a solid line, fuel injection is started when the nozzle needle 44 starts to be displaced (the fuel injection valve 40 is opened) with a delay of a predetermined amount Δ from the command injection start timing. In this case, as indicated by the one-dot chain line, if the actual valve opening start timing of the fuel injection valve 40 changes by the offset amount ΔTQ, the actual injection period shift does not become the offset amount ΔTQ. This is because, depending on the lift amount of the nozzle needle 44 when the command injection period elapses, the time from the elapse of time to the actual injection end changes.

  Even in this case, the deviation of the injection characteristic can be approximated by a straight line defined by the injection amount and the injection period shifted in the injection period direction. However, for example, after detecting the deviation of the actual injection amount, the offset amount ΔTQ, which is the deviation amount of the injection start timing, is quantified from the detected value and the command injection amount, and this may be used as the learning value. . Alternatively, the dependency of the injection amount deviation amount on the command injection amount may be learned by learning the deviation amount of the injection characteristic for each region divided by the injection amount and the fuel pressure. .

  Further, the deviation in the injection characteristic of the fuel injection valve 40 is not limited to that caused by the deviation in the delay amount of the actual injection start timing with respect to the command injection start timing. In any case, when performing multi-stage injection, it is effective to learn the amount of deviation in the injection characteristics by converting the difference between the predicted value and the detected value of the oxygen concentration into the difference due to the injection per time.

  -If processing for correcting the detection error of the sensor is performed separately, it is possible to learn the deviation of the injection characteristic of the fuel injection valve 40 with high accuracy by learning the learning value in a one-dimensional region divided by the fuel pressure. .

  In the second embodiment, instead of correcting the command injection amount with the learned value, the command injection period may be corrected with the learned value.

The figure which shows the whole structure of the engine system concerning 1st Embodiment. The flowchart which shows the process sequence of the fuel-injection control in the embodiment. The figure which shows the calculation map of request | requirement injection amount and target fuel pressure. The figure which shows the shift | offset | difference of the injection characteristic of a fuel injection valve. The time chart for demonstrating the factor of the shift | offset | difference of the injection characteristic of a fuel injection valve. The flowchart which shows the procedure of the process which learns the deviation | shift amount of the injection characteristic concerning the said embodiment. The figure which shows the map which memorize | stores the learning value in the same embodiment. 6 is a flowchart showing a processing procedure of oxygen concentration feedback control according to the embodiment; The flowchart which shows the process sequence of the fuel-injection control concerning 2nd Embodiment. The figure which shows the injection characteristic of the fuel injection valve concerning 3rd Embodiment. The figure for demonstrating the factor of the shift | offset | difference of the injection characteristic in the modification of each said embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Diesel engine, 40 ... Fuel injection valve, 70 ... Air flow meter, 76 ... Oxygen sensor, 90 ... ECU (one Embodiment of fuel-injection control apparatus), 92 ... Constant memory holding | maintenance memory (One embodiment of a memory | storage means).

Claims (9)

In a fuel injection control device for controlling the output of an internal combustion engine by multistage injection in which fuel injection is performed a plurality of times in one combustion cycle,
Means for calculating a predicted value of oxygen concentration in the exhaust gas of the internal combustion engine accompanying the fuel injection control;
Learning means for learning a deviation amount of a fuel injection valve of the internal combustion engine by converting a difference between the detected value of the oxygen concentration and the predicted value into a difference due to one injection of the multi-stage injection; A fuel injection control device comprising: The deviation of the injection characteristic is due to the deviation of the injection period due to the deviation of the delay amount of the actual injection start timing with respect to the command value of the injection start timing.
2. The fuel injection control apparatus according to claim 1, wherein the learning means learns the amount of deviation as a quantified amount of deviation in the injection period.   The learning means calculates a total deviation amount of the injection amount accompanying the multistage injection based on a difference between the detected value and the predicted value and an air amount to be burned in the internal combustion engine; 3. The fuel injection control device according to claim 1, further comprising a calculation unit that calculates the deviation amount of the injection characteristic by dividing the deviation amount by the number of injections of the multistage injection. 4. The fuel injection valve has a proportional relationship between the injection period and the injection amount, and has an inflection point at which the proportionality coefficient changes.
The learning means calculates a total deviation amount of the injection amount accompanying the multi-stage injection based on a difference between the detected value and the predicted value and an air amount provided for combustion in the internal combustion engine; A calculation means for calculating a deviation amount of the injection characteristic from the total deviation amount based on the number of injections of the multistage injection and the inflection point on the assumption that the injection period is shifted by a deviation amount of the delay amount of the injection start timing. The fuel injection control device according to claim 2, further comprising: The learning means learns the deviation amount of the injection characteristic for each of a plurality of regions defined by the pressure of fuel supplied to the fuel injection valve,
The fuel injection control device according to claim 1, further comprising storage means for storing a deviation amount of the injection characteristic learned for each of the plurality of regions. The learning means learns the amount of deviation in the injection characteristics for each of a plurality of regions defined by the pressure of the fuel supplied to the fuel injection valve and the rotational speed of the output shaft of the internal combustion engine,
The fuel injection control device according to claim 1, further comprising storage means for storing a deviation amount of the injection characteristic learned for each of the plurality of regions. Target value calculating means for calculating a target value of oxygen concentration in the exhaust;
Operation means for operating an actuator for controlling the oxygen concentration in the exhaust gas in order to feedback-control the predicted value to the target value;
The said operation means uses the predicted value calculated using the deviation | shift amount learned by the said learning means as the said predicted value used for the said feedback control, The Claim 1 characterized by the above-mentioned. Fuel injection control device. The internal combustion engine includes an exhaust gas recirculation passage for recirculating exhaust gas discharged to an exhaust system to an intake system, and an EGR valve for adjusting a flow area of the exhaust gas recirculation passage,
8. The fuel injection control apparatus according to claim 7, wherein the operating means operates an opening of the EGR valve so as to feedback control the predicted value to the target value. An opening / closing operation means for opening / closing the fuel injection valve based on a command value of an injection amount for the fuel injection valve;
The opening / closing operation means corrects the setting according to the deviation amount learned by the learning means when setting the operation amount of the fuel injection valve according to the command value. The fuel-injection control apparatus in any one of 1-6.

Images